Colorado, like most other states, currently is attempting to plan a continuum of treatment programs for the chronically mentally ill. These planning activities are hampered both by the lack of a unifying theory of treatment for this heterogeneous population and by a dearth of empirical verification for the planning models which are in place. The major goal of the proposed research is to empirically determine the types of chronically mentally ill individuals, and the optimum residential/service placement for each client type. The result of the research will be a generic state level planning and resource allocation model for the chronically mentally ill. In order to meet this goal, the proposed research has three major objectives. A. To develop a typology of the adult chronically mentally ill population. B. To select a random sample from each of the client types and subsequently collect data from these clients' primary therapists regarding the service and residential needs of these clients. The data will focus on five major areas. C. To analyze these data and produce a typology of residential/service mixes which characterize the needs of the various client types identified in Objective A. D. To disseminate the findings of the research in a manner which maximizes its utility for state level planners, policy and decision makers.